‘You shouldn’t always believe what you read in newspapers’

It is axiomatic that you shouldn’t always believe what you read in the newspaper. And so it is with a recent National Post story, headlined by this screamer: “78% favour abortions rights.”

The poll of just 608 people, conducted by Compass and commissioned by Global and the Post, was part of a larger poll on freedom of the body. The set of questions encompassed issues including doctor-assisted suicide (55 per cent in favour, 36 per cent against), freedom to join a legal sex club (“almost evenly split”), prostitution (52 per cent favour keeping it illegal) and smoking marijuana at home (52 per cent favoured the right).

Compass asked respondents: “Should women have complete freedom on their decision to have an abortion?” Without providing the percentage of respondents who said they do not support abortion or those who “did not know” or didn’t answer, the Post told readers that “four out of five Canadians, or 78%, believe women should have complete free choice in the matter.”

The Nov. 27 story was part of a larger series on liberty in Canada that featured poll results, articles and columns. Part of the problem, to be sure, is the need to educate Canadians on the fact that one right we should not have is to kill an unborn child in the womb, a fact acknowledged by Right to Life Toronto executive director Natalie Hudson: “It means we need to get out there more.”

But as Mary Ellen Douglas, national organizer for Campaign Life Coalition, told the Post, when a poll question focuses on the woman, the results tend to skew in favour of abortion. When the element of the child is brought in, however, polls show different results.

In fact, as the story acknowledged, the 78 per cent figure was 12 points higher than the previous high in an Environics poll from 2000.

But most other polls provide a more nuanced picture. As The Interim reported last month, a Leger Marketing poll of 3,010 people conducted in October found that a majority of Canadians oppose the status quo on abortion.

When asked from what point human life should be protected in law, only 30 per cent favoured it from the moment of birth. More than one-third, 37 per cent, favoured protecting life from the moment of conception, while 13 per cent favoured granting such protection at three months.

Another six per cent favoured legal protections beginning at six months and 14 per cent didn’t know or refused to answer. This proved Douglas’s point that answers are affected by whether a question on abortion focuses on the rights of the woman or the rights of the child.

The Leger Marketing poll is more in line with Gallup surveys of recent years. A Gallup poll conducted in December 2001 asked the question, “Do you think abortions should be legal under any circumstances, legal only under certain circumstances or illegal in all circumstances and in what circumstances?”

In that poll – which made no reference to the unborn – 32 per cent of Canadian respondents believed abortion should be legal in all circumstances, (down five points from the year before), 52 per cent thought abortion should be legal “only under certain circumstances” and 14 per cent thought abortions should be illegal in all circumstances. That means that 66 per cent of Canadians want at least some restrictions, a reversal of the Global/ Post/Compass findings.

(The Gallup poll also replicated Leger’s findings that many people take a gestational approach, with 57 per cent favouring abortion’s legality for the first three months.)

But the wording is not the only problem. Methodology presents its own limitations.

Writing in the Summer 2002 issue of Public Interest, Robert Weissberg noted the limitations of polling. He said, “The conventional poll” – that used by Compass in the Post story – “prohibits sub-optimal choices.” That is, there is no room for nuance or recognition of trade-offs. Such polls “create a world consisting only of first choices.”

This is especially true of the Post poll. The psychology of the respondents when asked a series of questions on freedom of the body will result in respondents being conditioned to favour generally libertarian answers. So, despite a personal view that may, for instance, favour abortion but with restrictions, the respondent will claim that he or she favours “complete freedom of choice.”